What Is Dark Social in B2B? Definition and Examples
A seller's definition of dark social, the private layer no tool can read and the public edge on LinkedIn you can watch, the SparkToro, Refine Labs and 6sense numbers behind it, a table against the dark funnel, direct traffic and intent data, and three actions for this week.
Dark social in B2B is the sharing and discussion of a vendor, a problem or a piece of content in private channels that analytics tools cannot see: Slack and Teams threads, WhatsApp groups, LinkedIn direct messages, forwarded emails, peer communities and calls. That is the short answer to "what is dark social in B2B". The conversations happen, they decide who gets on the shortlist, and when the buyer finally clicks through, your analytics file the visit under "direct traffic" with no source attached.
Where did the term dark social come from?
Alexis Madrigal coined "dark social" in October 2012 in The Atlantic, after looking at Chartbeat traffic data for the magazine and finding that a large share of visits arrived with no referrer at all, which meant the link had been pasted into an email, a chat or a text rather than clicked from a public feed (Wikipedia, Dark social media). The word "dark" means unseen, not sinister.
In B2B the term has stretched to cover every private place where buyers talk about a purchase: internal chat, peer groups, a podcast one person recommends to another, dinner at a conference. The common thread is that the conversation shapes the decision and leaves no trace you can attribute.
Why does dark social matter in B2B sales?
It matters because the decision is largely made before you know the buyer exists, and it is made in rooms you are not in. In 6sense's 2025 Buyer Experience Report, a survey of more than 4,000 buyers, 94% of buying groups had ranked a preferred vendor before first contact with any seller, and that vendor went on to win 80% of deals (6sense, 12 November 2025). That ranking is built from what colleagues say, what a peer group recommends and what gets forwarded, and none of it lands in a CRM.
The trace that does get left is misleading. Refine Labs compared what its attribution software recorded with what 620 buyers said over 12 months: software credited web search for 78% of conversions, buyers named it as their source 12% of the time, and 85% of the sources buyers named were social, podcasts, word of mouth and communities (Refine Labs, 15 October 2024). The sample is one company's, so read the ratio as an illustration rather than a law, but the direction matches what every team that asks the question reports.
A labelled example (hypothetical, not a customer): the head of sales at a 60-person staffing agency asks in a private Slack community of agency owners which tools they use to spot companies that are hiring. Four people answer. Two name the same product. She searches for it, reads the pricing page and books a demo. In that vendor's analytics she is "organic search". In its CRM she is an inbound lead. The thing that sold her, two recommendations in a channel of 300 peers, appears nowhere.
How does dark social work, and which part can a seller see?
Dark social works in two layers, and the split between them is the useful idea for a sales team.
The private layer is the thread itself: the Slack message, the WhatsApp group, the forwarded email, the call. No tool reads it, and you should not want one that does. The public layer is where those threads start and where they surface. Someone posts a question on LinkedIn. Someone comments under a competitor's launch. Someone shares a post with "this is exactly our problem". A private conversation usually begins with a public post that one person saw and forwarded, and it often ends with a public act: a follow, a like on your post, a connection request.
| Signal | Layer | Can you see it? | What to do |
|---|---|---|---|
| A buyer asks "anyone using a tool for X?" in a private community | Private | Only if you are a member | Join the two or three communities where your buyers talk, and answer when asked |
| A colleague forwards your post in Teams | Private | No; the visit arrives as direct traffic | Write posts worth forwarding; ask every new lead how they heard of you |
| A buyer posts the same question publicly on LinkedIn | Public | Yes | Reply with a real answer the same day, then connect |
| A buyer comments under a competitor's post | Public | Yes | Read what they wrote; message about the problem, never about the competitor |
| A buyer starts liking posts from three vendors in your category | Public | Yes, if you watch for it | Treat it as a shortlist in progress; engage before the demo request arrives |
| A new lead says "a friend recommended you" | Edge | Yes, if you ask | Record it in a required free-text field; thank the friend |
The private rows are why "dark social" exists as a term. The public rows are where a seller earns a place in the private thread. For those public rows, Intentsly watches LinkedIn for the posts, comments and competitor engagement that match your ideal customer profile and scores each person before you reach out, which is the daily watching in this table done for you; it does not and cannot read private channels.
Dark social vs dark funnel vs direct traffic vs intent data
These four terms get used as if they were interchangeable. They are not, and a seller usually needs to act on two of them at once.
| Dark social | Dark funnel | Direct traffic | Intent data | |
|---|---|---|---|---|
| What it names | Private sharing and discussion between people | Every buyer activity before your first recorded touch, private or public | An analytics bucket: visits with no referrer | Signals that an account or person is researching your category |
| Who uses the term | Marketers and sellers | Marketers, mostly | Analysts | Sales and sales operations |
| Is it a channel? | No, a description of behaviour | No, a phase of the purchase | No, a label | No, a type of data |
| Can you see it? | The public edge only | Parts of it | The count, never the source | Yes, that is its purpose |
| What to do | Earn forwards, ask the question, watch the public edge | Stop crediting the last click | Treat a spike as a sign something is being shared | Act on named people, not account scores |
Direct traffic deserves one more paragraph, because it is where dark social hides. Rand Fishkin's team at SparkToro ran a controlled test in April 2023, 1,113 visits over ten days from 16 tracked links: every visit from TikTok, Slack, Discord, Mastodon and WhatsApp arrived as "direct", 75% of Facebook Messenger visits did, and even 14% of visits from public LinkedIn posts lost their referrer (SparkToro, 27 April 2023). A seller who assumes "direct" means "typed our URL" is misreading the room. The lesson on intent data covers the account-level signals vendors sell to approximate the dark funnel, and where they stop short of a named person.
What we saw when we launched
When we launched Intentsly we had no way into anyone's Slack. We watched the public layer instead: founders posting on LinkedIn about cold outreach burnout, in their own words, and we let our AI SDR reply and hold the conversation. That produced 14 qualified demos in 14 days with zero manual outreach, and a reply rate 4x our previous cold campaigns (our case study). Most of those leads came from signals caught in real time, in public, from people whose private conversations we never saw and did not need to.
How to use dark social this week
Three actions, each under an hour.
- Add one question and make it required. Put a free-text "How did you hear about us?" on your demo form and ask it on every first call. Free text beats a dropdown because people write "a mate in a Slack group" rather than picking "Social". Expect memory bias, since people name the most recent or most memorable touch, so read the answers as direction rather than a ledger (Recast, 31 May 2023). After 30 answers you will know which communities and which people are sending you buyers.
- Join the rooms, and do not sell in them. Pick the two or three private communities (Slack, WhatsApp, a paid peer group) where your ideal customer profile actually talks. Answer questions fully, including when the right answer is a competitor. Being the person quoted in the thread is the only way into the private layer.
- Watch the public edge on LinkedIn every day. Save a search for the questions and pain phrases your buyers post, and look at who comments under your competitors' posts. The lesson on social listening in sales gives the phrase list, and the guide to monitoring competitors on LinkedIn in an hour a week gives the clicks. Reply the same day; a post that is going to be forwarded is usually forwarded within hours.
Related lessons
- What is a buying signal? defines the public acts in this lesson's table and gives a three-question test for which ones deserve a message.
- What is social listening in sales? is the practice of watching the public edge, with the phrases to listen for.
- What is social selling? is the long game that makes you the vendor people forward.
The rest of the Learn section defines the other terms used here. The free Buying Signal Playbook lists the 12 buying signals worth chasing and 8 to ignore, with a script for each. It is the shortest route from "we cannot see dark social" to a message you can send today.
FAQ
Is dark social the same as direct traffic?
No. Direct traffic is a bucket in your analytics for visits that arrive with no referrer. Dark social is one of the things that fills it, alongside typed URLs, bookmarks and some email clicks. SparkToro's 2023 test showed whole networks, Slack and WhatsApp among them, landing entirely in "direct", so a direct-traffic spike after a post usually means the post is being shared privately.
What is the difference between dark social and the dark funnel?
Dark social is the private sharing itself: the thread, the forward, the direct message. The dark funnel is the wider phase of a purchase that happens before your first recorded touch, which includes dark social but also public research you did not log, such as reading review sites or hearing a podcast. Dark social is behaviour; the dark funnel is a stage.
Can you measure dark social in B2B?
Not precisely, and anyone who quotes a single percentage for it is estimating. You can bound it: a required "how did you hear about us" field, win and loss interviews that ask the same question, and a look at direct traffic and branded search in the days after you publish. Combine what buyers tell you with what software records, because each one corrects the other's blind spot.
What are examples of dark social in B2B?
A LinkedIn post forwarded by direct message to a CFO. A case study pasted into a procurement team's email thread. A question in a private Slack community of agency owners. A WhatsApp group of marketing leaders swapping vendor names. A vendor mentioned on a podcast and discussed over lunch. In each case the discussion shapes the shortlist and leaves no trackable source.
Who coined the term dark social?
Alexis Madrigal, writing in The Atlantic in October 2012. He noticed in Chartbeat data that a large share of the site's visits arrived with no referrer and reasoned that the links were being shared in email, chat and text rather than on public social networks. The term later moved into B2B marketing to describe private buyer conversations.
What should a sales team do about dark social?
Stop trying to track it and start earning a place in it. Ask every lead how they heard of you, join the communities where your buyers ask for recommendations, and watch the public edge on LinkedIn, where private conversations begin and resurface. Reply to those public moments the same day, because that is what gets you forwarded.
